Related Experiment Video
Updated: Jun 30, 2026
Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
Long-term outcome prediction by clinicopathological risk classification algorithms in node-negative breast
Background:
Defining risk categories in breast cancer is of considerable clinical significance. We have developed a novel risk classification algorithm and compared its prognostic utility to the Web-based tool Adjuvant! and to the St Gallen risk classification.
Patients And Methods:
After a median follow-up of 10 years, we retrospectively analyzed 410 consecutive node-negative breast cancer patients who had not received adjuvant systemic therapy. High risk was defined by any of the following criteria: (i) age <35 years, (ii) grade 3, (iii) human epithelial growth factor receptor-2 positivity, (iv) vascular invasion, (v) progesterone receptor negativity, (vi) grade 2 tumors >2 cm. All patients were also characterized using Adjuvant! and the St Gallen 2007 risk categories. We analyzed disease-free survival (DFS) and overall survival (OS).
Results:
The Node-Negative-Breast Cancer-3 (NNBC-3) algorithm enlarged the low-risk group to 37% as compared with Adjuvant! (17%) and St Gallen (18%), respectively. In multivariate analysis, both Adjuvant! [P = 0.027, hazard ratio (HR) 3.81, 96% confidence interval (CI) 1.16-12.47] and the NNBC-3 risk classification (P = 0.049, HR 1.95, 95% CI 1.00-3.81) significantly predicted OS, but only the NNBC-3 algorithm retained its prognostic significance in multivariate analysis for DFS (P < 0.0005).
Conclusion:
The novel NNBC-3 risk algorithm is the only clinicopathological risk classification algorithm significantly predicting DFS as well as OS.